The purposes of this project are to find optimal methods of assessing physical activity energy expenditure (PAEE) in a free-living condition and to use these methods to determine the impacts of physical activity frequency, duration, and intensity on health, with particular emphasis on physical activity during pregnancy. We previously collected data, using a variety of accelerometers, heart rate monitors, and pedometers, with which we proposed to estimate PAEE in a cohort of healthy young adults. We simultaneously collected laboratory, nutrition, and anthropometric data from the same cohort. We have worked with the listed collaborators to develop and refine methods to analyze the data collected. We have also used the data derived from this study to develop methods for assessing time spent in sedentary behaviors. We have begun data collection in a trial to clarify the activity patterns, if any, that are most protective of both mother and fetus against the pregnancy complications attributed to obesity and diabetes. One member of our lab staff (JP) has analyzed data from Sweden and England regarding pregnant and non-pregnant women's activity patterns with an external collaborator (PWF). They found output from wrist-worn accelerometers explained at least 19% of variation in physical activity energy expenditure in non-pregnant women. They also found that maternal early insulin response was inversely correlated with physical activity during pregnancy, and that both of these variables were associated (early insulin response inversely and physical activity inversely) with infant fat-free mass. In a separate cohort, combined accelerometer--heart-rate-monitors, doubly labeled water, and indirect calorimetry were used to describe differences in activity subcomponents and energy expenditure between pregnant and non-pregnant women.